A question for the knowledgeable ones here.
Can I emulate, say, pulling the highlights slider to -20, by using the Tone curve? Is it a case of pulling the high end of the curve down, and perhaps adding an extra point to the left to limit the effect to the top end?
I guess what I’m getting at here, is whether those sliders are effectively standardised adjustments to the curve (presumably luma curve) or are they doing something different that the curve cannot achieve?
Yes, you can do that and results might even look close enough:
Looking at the whole situation reveals some differences, e.g. in the histopound.
Original image
Highlight slider
Tonecurve
Does it make sense to try to emulate? If you’re looking for some nerdy-geeky-task, it might be quite interesting to try to emulate the sliders, for practical reasons, I’d stick to the food that is on the menu.
Histopound?
A reminder to our friends in the US to vote wisely.
Can I refer to a thread that you and I participated in a while back…
This thread demonstrates the real problem with Selective Tone and why, if possible, you should try and use Tone Curve.
I’ve just created a comparison of a Tone Curve vs Selective Tone sliders, using the new Tone Curve points to accurately replicate the level of each step.
The top half is the Selective Tone sliders from the reference image and I have left the sliders visible but not active in the active image.
Basically, according to the ST sliders, I have supposedly…
My questions are…
…and we might add: What effect does such a (luma) curve have on colour’s?
@Joanna please, please be more specific. The problem with Selective Tones described in the post applies only to very deep blacks. With highlights there’s no such problem, i.e. excessive edge enhancement. If you state something like this, please mention the drawbacks of using the tone curves (e.g. bleached colors). Same goes with your opinions on ClearVision. Please stay responsible – too many people are listening to you. If you mean B&W, please make it explicit – it’s a different story; even I can tell that.
Then can you explain why there are edge enhancements between the rightmost (lighter tone) bands but, with these settings, not in the darker bands?
Here is another comparison…
Once again, the top half is the Selective Tone sliders from the reference image and I have left the sliders visible but not active in the active image.
According to the ST sliders, I have supposedly…
My questions this time are…
Here is the kind of curve I would expect to be the equivalent of changing the sliders like I have…
And you will notice there are no edge contrasts
Thank-you, Joanna. I remember similar presentations to address concerns that the PL selective tone sliders were not behaving like the Lr sliders. They still don’t. My take-away then was two-fold: stop comparing with Lr and stop using the PL selective tone sliders for global adjustments. The improved tone curve and local adjustments (which often include selective tone adjustments) are far better tools.
To keep true blacks truly black, obviously. If you like haze, you may use the tone curve. The zero value is shown in the 0-255, 8-bit scale, and it may correspond to non-zero in the 12, 14, or 16 bit raw data. It’s just another tool.
In some extreme cases they do, but mostly they make the image more readable. It mimics human perception better than LR version IMO, which often introduces oversaturated (negative blacks) or washed-out colors (positive blacks). But that’s entering the taste area.
In your last example you narrowed the output range between 72 and 255. No black under 72. I wonder if that is wanted.
if I have raised the blacks, how come the blackest black remains at 0?
Modification is within the range. The range itself isn’t altered.
George
Well, I’ve read this thread probably 20 times now… And I’m still confused.
What I understand:
The Selective Tone controls in PL do NOT operate like the ones in LR or C1. Okay, fair enough (I’m not sure if that’s a problem or not - I suspect not)
What I don’t understand:
Tom
I can appreciate your concerns, Tom. Having paid close attention to this issue for some time now, I am quite sure that:
(1) strictly speaking, the Selective Tone controls aren’t broken. They work as designed, for better or worse. DxO decided long ago to implement them differently than Adobe does and hasn’t changed their behavior even after lengthy and intense discussion. The addition of microcontrast, which sometimes negatively impacts gradients, isn’t unique to PhotoLab or to Selective Tone adjustments. Fortunately, the Tone Curve tool provides a way to make smoother, more gradual, and more bounded tonal adjustments. Local adjustments (the whole variety of them) are also useful in avoiding unwanted boundary/detail enhancement.
So my answer to your question (2) is: by all means, use them, with the understanding that they are not always the right tool for the job. 
I use global Selective Tone adjustment more often than not, but modestly. Same with ClearView Plus and Smart Lighting, which as has been noted can also either help or harm image quality. Most of my stronger adjustments are applied as local adjustments.
(3) I’d also like to have a tutorial, as I’ve rarely used the Tone Curve until now. I’m starting to experiment with it more in PL8. I understand the concept, but find it all too easy to make a mess with anything more than a very slight adjustment (for example, clipping highlights or crushing shadows or washing out part of the image).
Thank you, @Egregius. I appreciate your post.
Tom
With regards to the improved tone curve tool in PL 8, here are two recent videos:
As Greg has discussed, the selective tone controls are not broken, but they are idiosyncratic to PL. An extended discussion of the nuances of these sliders can be found here, but it is not an easy read (at least for me).
Interesting questions. Perhaps relevant, I found an old video that addresses similar slider vs RGB tone curve questions in Lr. There, the basic panel sliders apparently produce an even more exaggerated edge contrast and again no edge contrast using the tone curve. The plusses and minuses of each approach are discussed with a couple of examples.
